| Volume: 8; Issue: 2; December 2026 |
|---|
| Title : STATISTICAL POWER IN DATA-LIMITED CREEKS: SAMPLE SIZE AFFECTS t-TEST VS WILCOXON SENSITIVITY FOR ZOOPLANKTON IN AKASSA CREEK, DELTA STATE, NIGERIA |
| Authour(s): Efe Ogidìaka-Obende |
| Keywords: Akassa creek, Mann-Whitney U, Sample size, Statistical power, Zooplankton |
| Abstract: Data-limited coastal creeks in the Niger Delta are facing with limited baseline zooplankton data required for monitoring and management. A critical challenge is choosing the most appropriate statistical tests when data size is small and non-normal. This study compares statistical power of independent t-test and Mann-Whitney Wilcoxon Rank-Sum test for determining differences in zooplankton diversity and abundance across two stations in Akassa Creek. Data were collected in the year 2024. Statistical analysis was performed using Python 3 and the SciPy stats and pandas libraries for data manipulation and hypothesis testing. Four species were recorded in Akassa 1 while three species were collected from Akassa 2. The paired t-test showed no statistically significant difference in mean zooplankton abundance between Akassa 1 and Akassa 2 (t = 2.1267, p= 0.1234). Wilcoxon Signed-Rank Test showed no statistically significant difference in zooplankton abundance distribution between Akassa 1 and Akassa 2 (Test Statistic = 1.0000, p = 0.2500). With n = 4 taxa groups, the smallest possible p-value is 0.125 even if all ranks differ. The p = 0.2500 is therefore expected and does not prove 'no difference' but it proves 'test underpowered'. |
| DOI: https://doi.org/10.46505/IJBI.2026.8205 |
| How to cite this article: Ogidìaka-Obende Efe (2026). Statistical power in data-limited creeks: Sample size affects t-test vs Wilcoxon sensitivity for zooplankton in Akassa Creek, Delta State, Nigeria. International Journal of Biological Innovations. 8(2): 184-188. https://doi.org/10.46505/IJBI.2026.8205 |
Pages: 184-188 | Viewed: 6 | Downloaded: 5| Download: PDF (2864KB) |
